Oligonucleic acids are cleaved by a diverse range of antitumor antibiotics including the iron-bleomycin-oxygen complex and the enediynes. Each of these substances acts primarily by abstraction of a hydrogen atom from a C4' site on the target oligonucleotide. Experiments with authentically generated C4' radicals in single- stranded DNA have shown that their principle mode of decomposition is cleavage of the 3'O-P bond to give a radical cation, and that this cleavage is more rapid than quenching by oxygen. Yet, when the C4' radical is generated in double stranded DNA with bleomycin, trapping of the C4' radical by oxygen predominates. This proposal seeks to understand this difference in reactivity. It has been demonstrated, again with simple mononucleotide models, that the 2'O-substituent in RNA retards the fragmentation of C4' radicals very significantly. Thus, there is a pronounced substituent effect on the chemistry of C4' radicals. This proposal seeks to locate and then quantify the effect of the individual bases on the chemistry of C4' radicals at all levels. The effect of the base on hydrogen atom abstraction will be deterinined through use of the Norrish II photofragmentation sequence on appropriately functionallzed mononuleotides. Fragmentation of C4' radicals will be studied through unambiguous generation either by means of thiyl radical addition to exoglycals or by decarbonylation of a purpose-designed thiolester. Fragmentation kinetics will be determined under pseudo-first order conditions using a minor modification of a technique established in the PI's laboratory. The chemistry of the reaction of oxygen with C4' radicals, especially the possibility of its rever5ibility, will be investigated in detail. All experiments will be conducted with each of the four nucleosides usually found in DNA to asceriain the effect of the individual bases. Subsequently, the experiments will be repeated in the presence of incremental concentrations of matching purine or pyrimidine base with a view to understanding the influence of base pairing on the chemistry of the C4' radical. This is effectively a probe of the role of the second strand. Finally, a series of bicyclic C4' radical probes will be synthesized in order to probe a possible mechanical role of the second strand in retalding fragmentation of the C4' radical. The knowledge gained from this proposal, in concert with an understanding of factors governing binding selectivity, will enable new DNA cleaving antitumor antibiotics to be designed which target the most reactive nucleotides in double stranded DNA. This in turn will lead to increased efficiency of cleavage which ultimately translates to lower dosage and reduced toxcity to healthy cells.